FSA (version 0.8.30)

addZeroCatch: Adds zeros for catches of species not collected in some sampling events.

Description

Adds zeros for catches of species that were not captured in a sampling event but were captured in at least one other sampling event (i.e., adds zeros to the data.frame for capture events where a species was not observed).

Usage

addZeroCatch(df, eventvar, specvar, zerovar, na.rm = TRUE)

Arguments

df

A data.frame that contains the capture summary data as described in the details.

eventvar

A string for the variable that identifies unique capture events.

specvar

A string or vector of strings for the variable(s) that identify the “species” (if multiple variables, could be species, sex, and life stage, for example) captured. See examples.

zerovar

A string or vector of strings for the variable(s) that should be set equal to zero. See details and examples.

na.rm

A logical that indicates if rows where specvar that are NA should be removed after adding the zeros. See details.

Value

A data.frame with the same structure as df but with rows of zero observation data appended.

IFAR Chapter

2-Basic Data Manipulations

Details

The data.frame in df must contain a column that identifies a unique capture event (given in eventvar), a column with the name for the species captured (given in specvar), and a column that contains the number of that species captured (potentially given to zerovar; see details). All sampling event and species combinations where catch information does not exist is identified and a new data.frame that contains a zero for the catch for all of these combinations is created. This new data.frame is appended to the original data.frame to construct a data.frame that contains complete catch information -- i.e., including zeros for species in events where that species was not captured.

The data.frame may contain other information related to the catch, such as number of recaptured fish, number of fish released, etc. These additional variables can be included in zerovar so that zeros will be added to these variables as well (e.g., if the catch of the species is zero, then the number of recaptures must also be zero). All variables not given in eventvar, specvar, or zerovar will be assumed to be related to eventvar and specvar (e.g., date, gear type, and habitat) and, thus, will be repeated with these variables.

In situations where no fish were captured in some events, the df may contain rows that have a value for eventvar but not for specvar. These rows are important because zeros need to be added for each observed species for these events. However, in these situations, a <NA> species will appear in the resulting data.frame. It is unlikely that these “missing” species are needed so they will be removed if na.rm=TRUE (DEFAULT) is used.

One should test the results of this function by creating a frequency table of the eventvar or specvar. In either case, the table should contain the same value in each cell of the table. See the examples.

References

Ogle, D.H. 2016. Introductory Fisheries Analyses with R. Chapman & Hall/CRC, Boca Raton, FL.

See Also

complete in tidyr package.

Examples

# NOT RUN {
## Example Data #1 (some nets missing some fish, ancillary net data)
df1 <- data.frame(net=c(1,1,1,2,2,3),
                  eff=c(1,1,1,1,1,1),
                  species=c("BKT","LKT","RBT","BKT","LKT","RBT"),
                  catch=c(3,4,5,5,4,3))
df1
# not all 1s
xtabs(~net+species,data=df1)

df1mod1 <- addZeroCatch(df1,"net","species",zerovar="catch")
df1mod1
# check, should all be 3
xtabs(~net,data=df1mod1)
# check, should all be 1
xtabs(~net+species,data=df1mod1)
# correct mean/sd of catches
Summarize(catch~species,data=df1mod1)
# incorrect mean/sd of catches (no zeros)
Summarize(catch~species,data=df1)

# Same as example 1 but with no ancillary data specific to the net number
df2 <- df1[,-2]
df2
df1mod2 <- addZeroCatch(df2,"net","species",zerovar="catch")
df1mod2
# check, should all be 1
xtabs(~net+species,data=df1mod2)

## Example Data #3 (All nets have same species ... no zeros needed)
df3 <- data.frame(net=c(1,1,1,2,2,2,3,3,3),
                  eff=c(1,1,1,1,1,1,1,1,1),
                  species=c("BKT","LKT","RBT","BKT","LKT",
                            "RBT","BKT","LKT","RBT"),
                  catch=c(3,4,5,5,4,3,3,2,7))
df3
# should all be 1 for this example
xtabs(~net+species,data=df3)

# should receive a warning and table should still all be 1
df3mod1 <- addZeroCatch(df3,"net","species",zerovar="catch")
xtabs(~net+species,data=df3mod1)

## Example Data #4 (another variable that needs zeros)
df4 <- df1
df4$recaps <- c(0,0,0,1,2,1)
df4
# not all 1s
xtabs(~net+species,data=df4)

df4mod1 <- addZeroCatch(df4,"net","species",zerovar=c("catch","recaps"))
# note zeros in both variables
df4mod1
# check, should all be 1
xtabs(~net+species,data=df4mod1)
# observe difference from next
Summarize(catch~species,data=df4)
Summarize(catch~species,data=df4mod1)
# observe difference from next
Summarize(recaps~species,data=df4)
Summarize(recaps~species,data=df4mod1)

## Example Data #5 (two "specvar"s)
df5 <- df1
df5$sex <- c("m","m","f","m","f","f")
df5
# not all 1s
xtabs(~sex+species+net,data=df5)

df5mod1 <- addZeroCatch(df5,"net",c("species","sex"),zerovar="catch")
df5mod1
# all 1s
xtabs(~sex+species+net,data=df5mod1)
str(df5mod1) 

## Example Data #6 (three "specvar"s)
df6 <- df5
df6$size <- c("lrg","lrg","lrg","sm","lrg","sm")
df6

df6mod1 <- addZeroCatch(df6,"net",c("species","sex","size"),zerovar="catch")
df6mod1
 
# }